Method and system for using a bayesian belief network to ensure data integrity

• US 8,341,075 B2
• Filed: 09/13/2011
• Issued: 12/25/2012
• Est. Priority Date: 10/28/1999
• Status: Expired due to Term

• Current Assignee: Citibank (South Dakota) NA (Citigroup Incorporated)
• Original Assignee: Citibank (South Dakota) NA (Citigroup Incorporated)
• Inventors: Coleman, Ronald
• Primary Examiner(s): Akintola, Olabode
• Application Number:
  US13/231,240
• Publication Number:
  US 20120004949A1
• Time in Patent Office: 469 Days
• Field of Search: None
• US Class Current: 705/38
• CPC Class Codes: G06Q 30/0202 Market predictions or forecasting for commercial activities
• CPC Class Codes: G06Q 40/02 Banking, e.g. interest calculation or account maintenance credit or loans G06Q40/03
• CPC Class Codes: G06Q 40/03 Credit; Loans; Processing thereof
• CPC Class Codes: G06Q 40/04 Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
• CPC Class Codes: G06Q 40/06 Asset management; Financial planning or analysis
• CPC Class Codes: G06Q 40/08 Insurance

Abstract

The present invention relates to a method and system for assessing the risks and/or exposures associated with financial transactions using various statistical and probabilistic techniques. Specifically, the present invention relates to a method and system for identifying plausible sources of error in data used as input to financial risk assessment systems using Bayesian belief networks as a normative diagnostic tool to model relationships between and among inputs/outputs of the risk assessment system and other external factors.

Claims

1. A method for identifying plausible sources of error in a financial risk assessment (FRA) system, comprising:

identifying, by a computer, a plurality of variables of the FRA system;

implementing, by the computer, a Bayesian network to represent implications between and among the plurality of variables;

generating, by the computer, an initial probability for each of the plurality of variables of the FRA system that values for each of the plurality of variables have not changed;

extracting, by the computer, observed data from one of the plurality of variables of the FRA system;

determining, by the computer, a change in a value of at least one of the plurality of variables based on the extracted factual data from the FRA system; and

assessing, by the computer, the initial probability for the one of the plurality of variables of the FRA system based on the change in the value of one of the plurality of variables.

2. The method of claim 1, wherein the Bayesian network comprises a plurality of nodes corresponding to the plurality of variables.

3. The method of claim 2, further comprising:

assigning each one of the plurality of network nodes to one of the plurality of variables; and

assigning an initial probability to at least one of the plurality of network nodes.

4. The method of claim 1, wherein the plurality of variables comprise input data of the FRA system.

5. The method of claim 1, wherein one of the plurality of variables comprises information implicated from input data of the FRA system.

6. The method of claim 1, wherein the Bayesian network is implemented by a software having an applications program interface and a graphical user interface.

7. The method of claim 1, wherein generating the initial probability for each of the plurality of variables of the FRA system comprises:

setting each of the plurality of variables to a hypothesized state;

generating an initial probability for each of the plurality of variables in the set hypothesized state.

8. The method of claim 1, wherein the observed data comprise bias data and fact data about the one of the plurality of variables of the FRA system.

9. The method of claim 1, wherein extracting observed data from one of the plurality of variables of the FRA system comprises:

observing data from the one of the plurality of variables of the FRA system;

storing the observed data in a server archive; and

extracting the stored data out of the server archive.

10. A system for identifying plausible sources of error, the system comprising:

a financial risk assessment (FRA) system comprising a server, wherein the server is configured to;

identify a plurality of variables of the FRA system;

implement a Bayesian network to represent implications between and among the plurality of variables;

generate an initial probability for each of the plurality of variables of the FRA system that values for each of the plurality of variables have not changed;

extract observed data from one of the plurality of variables of the FRA system;

determine a change in a value of at least one of the plurality of variables based on the extracted factual data from the FRA system; and

assess the initial probability for the one of the plurality of variables of the FRA system based on the change in the value of one of the plurality of variables.

11. The system of claim 10, wherein the Bayesian network comprises a plurality of nodes corresponding to the plurality of variables.

12. The system of claim 10, wherein the plurality of variables comprise input data of the FRA system.

13. The system of claim 10, wherein one of the plurality of variables comprises information implicated from input data of the FRA system.

14. The system of claim 10, wherein the Bayesian network is implemented by a software having an applications program interface and a graphical user interface.

15. The system of claim 10, wherein the observed data comprise bias data and fact data about the one of the plurality of variables of the FRA system.

16. The system of claim 10, wherein extracting observed data from one of the plurality of variables of the FRA system comprises:

observing data from the one of the plurality of variables of the FRA system;

storing the observed data in a server archive; and

extracting the stored data out of the server archive.